To: Alan L.
From: Geoffrey Klempner
Subject: Wittgenstein's Tractatus: proposition 4.04
Date: 9 July 2003 12:49

Dear Alan,

Thank you for your e-mail of 30 June, with your first essay for the Philosophy of Language program, in response to the question:

'A gramophone record, the musical idea, the written notes, and the sound-waves, all stand to one another in the same internal relation to one another of depicting that holds between language and the world' ('Tractatus Logico-Philosophicus 4.04). - Discuss.

Lots of things to think about here.

You have shown considerable inventiveness in your use of the examples of a sound universe, and the graphical readout. But let's start with the idea that there is a fundamental *difference* between the language-world relationship (if we can call it that) and the relationship of 'depiction' which can hold between any two domains which exhibit a certain structural congruence, whether arising from a set of rules (as in written notes and sounds consequently produced by the orchestra) or a cause and effect relationship (grooves cut in the record in response to a sound input, which can then be used to generate the 'same' sound later).

Your explanation of the difference is that language is itself part of the world. You then explore this by considering what would happen if a musical score was itself 'part of the music', i.e. constructed out of sounds. Then you take this a step further by considering a universe consisting of only of sounds.

As it happens, there is a fascinating account of a 'sound world' in P.F. Strawson 'Individuals: an essay in descriptive metaphysics' Ch. 2. Strawson is concerned with the possibility of distinguishing between a sound-world subject's 'subjective' stream of impressions, and the 'objective' arrangement of sounds, and he raises some difficulties of principle (which actually I think can be surmounted). The aim is to exhibit the minimal experiential structure required for having a conception of oneself as a subject in relation to a world of objects. But let's suppose we could explain the distinction between experience and its objects in a sound world. Then I see no reason why there could not be a language consisting of sounds, which, as they were produced, became objects in the sound world, just as human utterances or writings are objects in our world.

Here is my explanation of the difference between language and other forms of 'depiction'. This will tie in with your claim that language 'summarises and abbreviates the world'.

The limits of musical notation, or any kind of non-linguistic depiction are determined by a circumscribed 'object' in the world - a musical performance, a geographical terrain, a machine-woven cloth. Language, by contrast, is the one and only species of depiction which is limited to 'whatever may (or may not) be the case'.

I look up at the sky and say, 'It is cloudy today.' That is a lot less than all that can be said, of course. In another sense, my statement is completely adequate, just as it stands. (It would be inadequate if 'being cloudy' was not a coherent concept.) Wittgenstein is insistent that the world IS all that is the case, and all that is the case cannot, in principle, go beyond all that can be truly said. That is consistent with the realization that I could die of old age before I had said *everything* that can be said about the cloudy sky.

A musical score, an ordinance survey map, a cloth weaving template are not 'propositions' of language, although they function in a manner in some ways analogous to language.

Your observations about the graphical read-out illustrate the nature of logic, and its connection with a given system of linguistic signs very elegantly. Wittgenstein at one point in the Tractatus considers the intriguing question of possibilities of different 'size of mesh' in his observations about Newtonian mechanics (6.341, 6.342). One might puzzle, though, about Wittgenstein's assertion that "What *does* characterize the picture is that it can be described *completely* by a *particular* size of mesh." If the 'mesh' we choose for our language determines what aspects of the world will become visible to us, how do we ever discover that the mesh is too coarse? Not by somehow bypassing language and making direct contact with the world. So what exactly happens when we come across a phenomenon that the existing language is not adequate describe?

Let's say that I am observing a graphical readout and need to communicate the results to you. I decide on one system of representation which seems to work well, but then run into problems because I encounter situations which my system is inadequate to describe. But this is a limited system of depiction, not language as such.

The case of your 'gentle curve' illustrates another difficulty. However fine the discrimination, there will still be borderline cases. It's not like, e.g. writing a notation for harpsichord music and then needing a new device for 'loud' and 'soft' when someone invents the piano. As I think you have realized, the question of vagueness is the Achilles heel of the whole Tractarian enterprise.

All the best,

Geoffrey